# Marc Andreessen says OpenAI's medical AI makes fewer errors than human doctors

> Marc Andreessen claims AI outperforms 99.99% of doctors after OpenAI's GPT-5.6 received fewer flaw ratings. U.S. hospitals are deploying these tools.

- Type: AI news
- Page: https://completeaitraining.com/news/marc-andreessen-says-openais-medical-ai-makes-fewer-errors/
- Published: 2026-07-12
- Updated: 2026-07-12
- Author: Jeroen Erne
- Jobs: Healthcare
- Reading time: 1 min

Venture capitalist Marc Andreessen said artificial intelligence already outperforms nearly all physicians after OpenAI reported that its latest model, GPT-5.6, produced medical responses clinicians rated less flawed than answers written by human doctors. The claim, posted on social media, arrives as health systems across the U.S. accelerate deployment of AI-powered patient tools.

Andreessen was reacting to an update from OpenAI CEO Sam Altman. "Physicians found fewer flaws in GPT-5.6 responses than physician-written responses," Altman said. Andreessen replied, "AI is already a better doctor than 99.99% of human doctors. This is good news."

#### How health systems are embedding AI

Major providers are weaving AI chatbots into patient-facing workflows. Hartford HealthCare launched Patient GPT from K Health, while Sutter Health and Reid Health rolled out Epic's Emmie platform. The tools operate inside HIPAA-protected environments, using electronic health records to support scheduling, medical information delivery, and physician tasks.

The Mayo Clinic partnered with Microsoft to blend clinical expertise and de-identified patient data with the company's AI and cloud capabilities, aiming to sharpen treatment decisions and outcomes. Cathie Wood, CEO of ARK Invest, said AI has a "profound application" in healthcare, pointing to its potential to speed medical research, diagnosis, and drug development. Healthcare professionals seeking to build skills around these tools can explore [AI for Healthcare Courses](https://completeaitraining.com/tag/healthcare/) designed for clinical adoption.

#### Why this matters for healthcare professionals

The GPT-5.6 finding does not mean AI can replace physician judgment. Real-world diagnosis relies on nuanced physical exams, patient histories, and ethical reasoning that language models cannot duplicate. The fast-paced rollout of chatbots across health systems puts a premium on clinicians who critically evaluate AI outputs and retain full accountability for care decisions. Understanding where these tools misfire - and where they genuinely assist - becomes a core skill for anyone practicing medicine alongside AI.

---

This is the Markdown form of https://completeaitraining.com/news/marc-andreessen-says-openais-medical-ai-makes-fewer-errors/. The whole site is mapped at https://completeaitraining.com/llms.txt.